计算机科学

计算机科学 ›› 2026, Vol. 53 ›› Issue (6A): 250400146-11.doi: 10.11896/jsjkx.250400146

• 大数据&数据科学 • 上一篇    下一篇

基于TransLSTM-GAN模型的碳排放预测算法

张筱竹1, 陈泓佑1, 屈凌峰2, 王玥陈佳1, 田宝单3, 范勇1   

  1. 1 西南科技大学四川省大数据与智能系统工程技术研究中心 四川 绵阳 621010
    2 广州大学网络空间先进技术研究院 广州 510006
    3 西南科技大学数理学院 四川 绵阳 621010
  • 出版日期:2026-06-16 发布日期:2026-06-12
  • 通讯作者: 陈泓佑(chy2019@foxmail.com)
  • 作者简介:(XiaozhuZhang2025@163.com)
  • 基金资助:
    四川省科技计划(2025ZNSFSC0005);国家自然科学基金(62402125)

Carbon Emission Prediction Algorithm Based on TransLSTM-GAN Model

ZHANG Xiaozhu1, CHEN Hongyou1, QU Lingfeng2, WANG Yuechenjia1, TIAN Baodan3, FAN Yong1   

  1. 1 Sichuan Big Data and Intelligent System Engineering Technology Research Center,Southwest University of Science and Technology,Mianyang,Sichuan 621010,China
    2 Cyberspace Institute of Advanced Technology,Guangzhou University,Guangzhou 510006,China
    3 School of Mathematics and Physics,Southwest University of Science and Technology,Mianyang,Sichuan 621010,China
  • Published:2026-06-16 Online:2026-06-12
  • About author:ZHANG Xiaozhu,born in 2004,undergraduate,is a member of CCF(No.N1031G).Her main research interests include deep learning and image processing.
    CHEN Hongyou,born in 1989,Ph.D,lecturer,master's supervisor.His main research interests include deep learning,image processing,intelligent computational fluid dynamics and information security.
  • Supported by:
    Sichuan Science and Technology Program (2025ZNSFSC0005) and National Natural Science Foundation of China(62402125).

PDF (PC)

60

摘要： 碳排放预测对国际合作、应对气候变化和能源安全等极为重要。国家级地理范围和更长时间段内的碳排放预测影响因素多而复杂,这对预测模型的特征表达学习能力提出了更高要求。针对以上问题,提出了一种融合Transformer,LSTM和GAN的碳排放预测模型——TransLSTM-GAN。在这项工作中,通过Transformer和LSTM网络,利用注意力机制的高性能特征表达学习能力,提高模型处理复杂碳排放数据和长序列数据的学习能力。设计适配的改进鲸鱼优化算法,完成Trans-LSTM 的超参数自动学习,降低TransLSTM的训练难度和提升训练效果。利用预训练TransLSTM作为生成器,深度残差网络作为判别器,构建GAN对生成器参数微调,进一步提高预测精度。为验证模型性能,在中国、美国和欧洲碳排放数据集上进行实验验证。实验结果表明,TransLSTM-GAN模型能更好地适用于不同国家地区的长期碳排放预测。

关键词: TransLSTM, 生成式对抗网络, 深度学习, 鲸鱼优化算法, 碳排放预测

Abstract: Carbon emission prediction is crucial for several aspects,including international cooperation,addressing climate change,and energy security.Due to the numerous and complex factors affecting carbon emission prediction within national geographic area and over longer time interval,there are higher requirements for the feature representation learning ability of prediction mo-dels.Aiming to the above problems,a carbon emission prediction model that integrates transformer,long short-term memory(LSTM) neural network,and generative adversarial network(GAN) is proposed,called TransLSTM-GAN.In this work,attention mechanism and high-performance feature representation learning ability are utilized via Transformer and LSTM network to improve the model learning ability in processing complex carbon emission data and long sequence data.An adaptive improved whale optimization algorithm(IWOA) is designed to automatic hyperparameter learning for TransLSTM,reducing training difficulty and improving training effectiveness.Using pre-trained TransLSTM as a generator and deep residual network(ResNet) as a discriminator,a GAN is constructed to fine tune the generator parameters and further improve prediction accuracy.To validate the performance of this model,experimental verification on carbon emission datasets in China,America,and Europe.The experimental results indicate that the TransLSTM-GAN model can better adapt to national regions and predict long-term carbon emissions.

Key words: TransLSTM, GAN, Deep learning, Whale optimization algorithm, Carbon emission prediction

中图分类号: 

  • TP391
[1] KUMAR S,SHUKLA A K,MUHURI P K.Anomaly based novel multi-source unsupervised transfer learning approach for carbon emission centric GDP prediction[J].Computers in Industry,2021,126(2021) 103396:1-12.
[2] DESAI A,GANDHI S,GUPTA S,et al.Carbon emission prediction on the world bank dataset for Canada[J].arXiv:2211.17010,2022.
[3] CHU X L,ZHAO R J.A building carbon emission prediction model by PSO-SVR method under multi-criteria evaluation[J].Journal of Intelligent and Fuzzy Systems,2021,41(6):7473-7484.
[4] SONG Y C,SHU Q,JIN C R,et al.Regional transport carbon emission forecasting and peak carbon pathway planning in China[J].Environmental Science,2025,46(4):1995-2008.
[5] ZHAO X N,XIE X L,ZHAO R J.A method for predicting carbon emission of railway transportation system based on an LSTM network with dynamic input via Sliding window[J].Journal of Transport Information and Safety,2023,41(1):169-178.
[6] SHAO C H,NING J.Construction and application of carbon emission prediction model for China's textile and garment industry based on Improved WOA-LSTM[J].Journal of Beijing Institute of Fashion Technology(Natural Science Edition),2023,43(4):73-81.
[7] ZHANG X M,WANG X X,CHEN Y H.Carbon emission prediction and clean industry transformation based on machine learning:A case study of Sichuan province[J].arXiv:2309.01115,2023.
[8] WANG P X,JING Z,ZHANG Z,et al.CasieNet:Category-aware and side information enhanced pre-training network for carbon emission prediction[C]//International Conference on Innovation in Artificial Intelligence(ICIAI).2023:79-85.
[9] LI Y J,YANG P,ZHAO S Y,et al.Development trend capture and short-term forecasting for carbon emission based on VMD-RE-LSTM[J].Environmental Engineering,2023,41(S2):279-285,249.
[10] QIN H Y,HAN M.Real-time forecasting and short-term prediction of Chinese carbon dioxide emissions based on MIDAS model[J].Acta Scientiae Circumstantiae,2018,38(5):2099-2107.
[11] WANG Y,TANG Z H,WANG Z X,et al.Short-term forecasting of carbon emission by using improved Gaussian process regression[J].Computer Engineering and Applications,2018,54(23):246-251,258.
[12] ZHANG K Y,MENG T N,LIU R J,et al.Design of short-term carbon emission prediction model based on secondary decomposition strategy and BiLSTM[J].Electronic Design Engineering,2024,32(17):6-10.
[13] HE X,ZHANG Q S,WANG Y D,et al.Research on PISC-VMD-GRU carbon emission prediction algorithm[C]//International Conference on Artificial Intelligence,Big Data and Algorithms(CAIBDA).Zhengzhou,China,2024:902-907.
[14] WANG Y L,LI Y W,WANG H,et al.Carbon Emission Prediction of Hebei Province Based on Improved BP Neural Network[J].Ecological Economy,2024,40(06):30-37.
[15] LIAN Y Q,SU D H,SHI S X.Carbon peak prediction in Fujian province based on combined STIRPAT and CNNLSTM models[J].Environmental Science,2025,46(1):10-18.
[16] LI G Z,HUANG Q H.Prediction of carbon peak in Beijing-Tianjin-Hebei region based on Lasso-GRNN neural network model[J].Environmental Science,2025,46(2):636-646.
[17] WANG W J,PAN H,WANG G G.Prediction of industrial carbon peaking in Liaoning province based on GWO-LSTM model[J].Environmental Science and Management,2024,49(1):28-33.
[18] HUANG W,WEI D Q,WANG C,et al.Matrix completion-based prediction analysis in carbon emissions[J].International Journal of Embedded Systems,2021,14(2):143-148.
[19] TANG J H,FANG N,YANG L Q,et al.CarbonNet:Enter-prise-level carbon emission prediction with large-scale datasets[C]//Advanced Intelligent Computing Technology and Applications International Conference(ICIC).Tianjin,China:2024,411-422.
[20] WU X C,FENG X,JIANG H.Research on carbon emission prediction based on multi head attention CNN-LSTM[J/OL].Journal of Chongqing Technology and Business University(Natural Science Edition),2024.
[21] WANG H P,ZHANG Z.Forecasting Chinese provincial carbon emissions using a novel grey prediction model considering spatial correlation[J].Expert Systems with Applications,2022,209(2022)118261:1-19.
[22] ZHANG D B,YU Z H,LING L W,et al.A combined framework for carbon emissions prediction integrating online search attention[J].Journal of Intelligence and Fuzzy Systems,2024,46(4):11153-11168.
[23] BOSAH P C,IBRAHIM A,LI S X,et al.Predicting carbon dioxide emissions in the United States of America using machine learning algorithms[J].Environmental Science and Pollution Research,2024,31:33685-33707.
[24] YU Y,SI X S,HU C H,et al.A review of recurrent neural networks:LSTM cells and network architectures[J].Neural Computation,2019,31(7):1235-1270.
[25] VASWANI A,SHAZEER N,PARMAR N,et al.Attention isall you need[C]//Advances in Neural Information Processing Systems(NeurIPS).Long Beach,USA:2017:5998-6008.
[26] ZHANG H Q,SONG H L,LI S Y,et al.A survey of controllable text generation using transformer-based pre-trained language models[J].ACM Computing Survey,2024,56(3):1-37.
[27] BOZORGI S M,YAZDANI S.IWOA:An improved whale optimization algorithm for optimization problems[J].Journal of Computational Design and Engineering,2019,6:243-259.
[28] CHENAN H,TSUTSUMIDA N.A scalable k-medoids cluste-ring via whale optimization algorithm[J].arXiv:2408.16993,2024.
[29] OUYANG C T,GONG Y K,ZHU D L,et al.Improved whale optimization algorithm based on fusion gravity balance[J].Axioms,2023,12(7):1-30.
[30] ZHANG J Z,KONG M,ZHANG G,et al.Weapon-target assignment using a whale optimization algorithm[J].International Journal of Computational Intelligent Systems,2023,16(1):1-18.
[31] GOODFELLOW I J,POUGET-ABADIE J,MIRZA M,et al.Generative adversarial nets[C]//Advances in Neural Information Processing Systems(NeurIPS).Montreal,Canada:2014:2672-2680.
[32] CHEN H Y,HE H J,CHEN F,et al.A training method of DCGANs based on subsample set construction[J].ACTA AUTOMATICA SINICA,2021,47(4):913-923.
[33] CHEN H Y,CHEN F,HE H J,et al.A GANs model based on sample feature decoding constraint[J].ACTA AUTOMATICA SINICA,2022,48(9):2288-2300.
[34] HE K M,ZHANG X Y,REN S Q,et al.Deep residual learning for image recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Las Vegas,USA:2016,770-778.
[35] HE K M,ZHANG X Y,REN S Q,et al.Identity mappings indeep residual networks[C]//European Conference on Computer Vision(ECCV).Amsterdam,The Netherlands:2016:630-645.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市两江新区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发